AI4RTC – AI for Real Time Charging

Powered by   

Local AI along with EV Loader have been awarded the AI4RTC – AI for Real Time Charging projectand were funded by the Horizon2020 TrialsNet project to develop and trial an innovativesolution to perform real-time Load Management of EV Charging Stations.

In this project we take advantage of the AI algorithms of Local AI for Timeseries Forecasting ofthe power consumption of EV Chargers and develop a platform that can communicate with theCPMS of a live EV Charging Network (EV Loader) over the standard protocol OCPI and preventpower spikes that can stress the Grid and harm the stability of the local power network.

During the duration of the project we have conducted multi-month active trials on the 5 EVLoader charging sites.

We have exploited the low latency of the #Cosmote 5G SA network to achieve real-time #AIbased load-management.

This project has received the valuable support of the TrialsNet consortium and we had greatcollaboration with our partner EV Loader.

We aim to provide a greater stability to both the Grid and to the fast-growing EV chargingnetworks in Europe.

We have selected 5 Sites in the Greater Athens area where we have installed 5G-SA connectivity infrastructure, and we have tested the performance of Cosmote 5G-SA network KPIs.

Local AI has tested extensively the accuracy of the AI timeseries forecasting algorithms that use state of the art techniques such as Attention Layers (Transformers) and TCN Layers (Temporal Convolutional Network)

The results we very satisfactory and we were able to capitalise on them to achieve real time power control for the pilot sites.